The issue might be due these reasons:-

Model Training Data: spaCy's pre-trained models are trained on specific datasets. If certain entities or terms were not sufficiently represented in the training data, the model might not recognize them as entities.
Model Limitations: Every model has its limitations. The pre-trained models may not always capture all entities accurately.
Language Model:- The performance of entity recognition can vary between different language models. For example, the es_core_news_lg and it_core_news_lg models are specifically trained for Spanish and Italian, respectively. If the entities you are trying to extract are domain-specific or less common, these models might not perform well.

To solve the issue you may try these steps and let me know, if it works::-

Custom Training Train a custom Named Entity Recognition (NER) model with your specific dataset.
Data Augmentation If you have a small dataset, consider augmenting it with more examples or using transfer learning
Entity Ruler Use spaCy's EntityRuler to add rules-based entity extraction.
Model Evaluation and Fine-Tuning Evaluate the performance of different spaCy models and fine-tune them to better suit your needs. sample code:-
```
import spacy
from spacy.pipeline import EntityRuler
```

Load the spaCy model

nlp = spacy.load("es_core_news_lg") # or "it_core_news_lg"

Create an EntityRuler and add patterns

ruler = EntityRuler(nlp, overwrite_ents=True) patterns = [ {"label": "ORG", "pattern": "OpenAI"}, {"label": "PRODUCT", "pattern": "ChatGPT"},

] ruler.add_patterns(patterns)

nlp.add_pipe(ruler, before="ner")

doc = nlp("OpenAI has developed ChatGPT.")

for ent in doc.ents: print(ent.text, ent.label_)


Hope this helps, Thanks